Home | History | Annotate | Download | only in tools 
      1 #!/usr/bin/python
      2 # Run with directory arguments from any directory, with no special setup required.
      3 # Or:
      4 # for i in libc libdl libm linker libstdc++ libthread_db ; do ./libc/tools/generate-NOTICE.py $i > $i/NOTICE ; done
      5 
      6 import ftplib
      7 import hashlib
      8 import os
      9 import re
     10 import shutil
     11 import string
     12 import subprocess
     13 import sys
     14 import tarfile
     15 import tempfile
     16 
     17 def IsUninteresting(path):
     18     path = path.lower()
     19     if path.endswith(".mk") or path.endswith(".py") or path.endswith(".pyc") or path.endswith(".txt") or path.endswith(".3"):
     20         return True
     21     if path.endswith("/notice") or path.endswith("/readme") or path.endswith("/caveats"):
     22         return True
     23     if path.endswith("/tzdata") or path.endswith("/zoneinfo/generate"):
     24         return True
     25     return False
     26 
     27 def IsAutoGenerated(content):
     28     if "generated by gensyscalls.py" in content or "generated by genserv.py" in content:
     29         return True
     30     if "This header was automatically generated from a Linux kernel header" in content:
     31         return True
     32     return False
     33 
     34 copyrights = set()
     35 
     36 def ExtractCopyrightAt(lines, i):
     37     hash = lines[i].startswith("#")
     38 
     39     # Do we need to back up to find the start of the copyright header?
     40     start = i
     41     if not hash:
     42         while start > 0:
     43             if "/*" in lines[start - 1]:
     44                 break
     45             start -= 1
     46 
     47     # Read comment lines until we hit something that terminates a
     48     # copyright header.
     49     while i < len(lines):
     50         if "*/" in lines[i]:
     51             break
     52         if hash and len(lines[i]) == 0:
     53             break
     54         if "\t@(#)" in lines[i] or "\tfrom: @(#)" in lines[i] or "From: @(#)" in lines[i] or "from OpenBSD:" in lines[i]:
     55             break
     56         if "\tcitrus Id: " in lines[i]:
     57             break
     58         if "\t$OpenBSD: " in lines[i] or " $FreeBSD: " in lines[i] or "\t$NetBSD: " in lines[i]:
     59             break
     60         if "$FreeBSD$" in lines[i] or "$Citrus$" in lines[i]:
     61             break
     62         i += 1
     63 
     64     end = i
     65 
     66     # Trim trailing cruft.
     67     while end > 0:
     68         if lines[end - 1] != " *" and lines[end - 1] != " * ====================================================":
     69             break
     70         end -= 1
     71 
     72     # Remove C/assembler comment formatting, pulling out just the text.
     73     clean_lines = []
     74     for line in lines[start:end]:
     75         line = line.replace("\t", "    ")
     76         line = line.replace("/* ", "")
     77         line = line.replace(" * ", "")
     78         line = line.replace("** ", "")
     79         line = line.replace("# ", "")
     80         if line.startswith("++Copyright++"):
     81             continue
     82         line = line.replace("--Copyright--", "")
     83         line = line.rstrip()
     84         # These come last and take care of "blank" comment lines.
     85         if line == "#" or line == " *" or line == "**" or line == "-":
     86             line = ""
     87         clean_lines.append(line)
     88 
     89     # Trim blank lines from head and tail.
     90     while clean_lines[0] == "":
     91         clean_lines = clean_lines[1:]
     92     while clean_lines[len(clean_lines) - 1] == "":
     93         clean_lines = clean_lines[0:(len(clean_lines) - 1)]
     94 
     95     copyright = "\n".join(clean_lines)
     96     copyrights.add(copyright)
     97 
     98     return i
     99 
    100 args = sys.argv[1:]
    101 if len(args) == 0:
    102     args = [ "." ]
    103 
    104 for arg in args:
    105     sys.stderr.write('Searching for source files in "%s"...\n' % arg)
    106 
    107     for directory, sub_directories, filenames in os.walk(arg):
    108         if ".git" in sub_directories:
    109             sub_directories.remove(".git")
    110         sub_directories = sorted(sub_directories)
    111 
    112         for filename in sorted(filenames):
    113             path = os.path.join(directory, filename)
    114             if IsUninteresting(path):
    115                 #print "ignoring uninteresting file %s" % path
    116                 continue
    117 
    118             try:
    119                 content = open(path, 'r').read().decode('utf-8')
    120             except:
    121                 # TODO: update hash.h, md5.c, and md5.h; upstream is probably UTF-8 already.
    122                 sys.stderr.write('warning: bad UTF-8 in %s\n' % path)
    123                 content = open(path, 'r').read().decode('iso-8859-1')
    124 
    125             lines = content.split("\n")
    126 
    127             if len(lines) <= 4:
    128                 #print "ignoring short file %s" % path
    129                 continue
    130 
    131             if IsAutoGenerated(content):
    132                 #print "ignoring auto-generated file %s" % path
    133                 continue
    134 
    135             if not "Copyright" in content:
    136                 if "public domain" in content.lower():
    137                     #print "ignoring public domain file %s" % path
    138                     continue
    139                 sys.stderr.write('warning: no copyright notice found in "%s" (%d lines)\n' % (path, len(lines)))
    140                 continue
    141 
    142             i = 0
    143             while i < len(lines):
    144                 if "Copyright" in lines[i]:
    145                     i = ExtractCopyrightAt(lines, i)
    146                 i += 1
    147 
    148             #print path
    149 
    150 for copyright in sorted(copyrights):
    151     print copyright.encode('utf-8')
    152     print
    153     print '-------------------------------------------------------------------'
    154     print
    155 
    156 sys.exit(0)
    157